My last week at the NVIDIA GTC 2024 conference was essentially a week’s worth of 🤯🤯🤯mind-blown emoji’s🤯🤯🤯 . Everyone will have their different take, based on what they were able to see during their time there. Here's a breakdown of my top 10 insights of what struck me the most, minus the tech jargon:
1. AI is the New Electricity⚡: It will be as ubiquitous as electricity is today, and become a utility that simply powers every single thing. The foundational model chatbots will be the power plugs in our houses that connects us to all AI capabilities.
2. Humanizing the AI: Imagine walking into a store and having a friendly employee answering your questions. Now ecommerce sites can replicate that experience with human-like AI avatars that can do the same. This can lead to a whopping 5x increase in conversion rates! The impact isn’t coming so much from better information about the products, but more from the human need of validation that they’re making the right choice.
3. Digital Twins Help Manufacturing and Supply Chain Leap into AI: I tend to think of manufacturing and supply chain as old-school physical world based industries that will be very late to the game with AI. However, digital twins (full virtual replicas of facilities and warehouses) allow AI to learn, and optimize endlessly BEFORE deploying in real-life. A full digital twin of a chip manufacturing facility was able to not only reduce power consumption by 30% but shortened build time by half because of all the optimizations pre-run in the digital twin.
4. Forget Traffic Jams, Hello Self-Driving Cities: It’s not just self-driving cars anymore. You have self-driving buses for public transport, self-driving vendor carts selling concession, self-driving trailer trucks for long-haul transportation, and fully autonomous child-care and extracurricular pick ups and drop offs. Ok, I added that last one myself, but I have 3 teenagers with 9 extra curricular activities among them, can you blame me?
5. Design Your Dream Home (Virtually) and Buy It: The Apple Vision Pro can allow you collaborate with an interior designer on the living room of your dreams and buy everything when you’re done. All the furniture will fit because it’s designed in a digital twin of your house. For all of us that buy furniture online and spend hours assembling before realizing that it’s too big for our space, where do I get this today?
6. Jensen Huang* is the Taylor Swift of Tech: I literally heard the comparison 3 times in 2 days. I don’t think there’s any resemblance, but the impact might be equivalent. For my choice of a pop star, I would definitely wait in line for Jensen tickets. (*Jensen Huang is the CEO of NVIDIA)
7. Unleashing Creativity with AI: Real-time 3D AI art opened the NVIDIA keynote using generative AI models of nature. I love that the graphics DNA of NVIDIA brought the creative potential to the foreground for AI. Creators also showcased fantastical immersive experiences that were adjusted in real-time to meet the audience’s preferences. Creator tools are also making the leap to create full 3D content easily for today’s tiktok content creators. If anyone can easily create 3D fantasy worlds that others can experience, what new creative art forms will emerge?
8. The Rise of the Citizen Scientist: AI is making complex scientific data analysis accessible to everyone. Citizen scientists can contribute to research efforts, leading to faster breakthroughs. This is one of my favorites because I’m too lazy to actually do a PhD program, but I want to ask the big questions and experiment with big ideas. While I think my application to a researcher position at NVIDIA has a low likelihood of being accepted, I’m happy to settle for making my own small research projects for fun.
9. Robots That Learn on the Fly: Forget Complex Coding! AI robots are becoming so sophisticated that they can learn to perform new tasks simply by watching demonstrations. This is 100 miles from me having to code axel rotations to get little robot to dance one move. I can see a robot in my home when I retire doing things like watering the plants regularly, and taking the dog on a walk in the rain. This will happen in my lifetime.
10. Language is Power, but There’s a New Vocabulary: At the beginning of the ChatGPT wave, I could see that the natural language chat interface was a great unlocking of being able to direct technology to complete tasks, write, create, and interact with the real world. But there’s a New Vocabulary that will allow some users to unlock more AI power than others: words like reinforcement learning, self reflection, and orchestration. The best AI power users of tomorrow will have mastered the nuances of this vocabulary and knows exactly when to use them in their prompts to achieve phenomenal abilities from their AI. How do we teach that to our kids?
I begin my career at the dawn of the internet, then came mobile, and social media. The amount of data and content generated from those 3 technologies is pushing us into the golden age of AI. I will see this one as well in my life time, and it’s mind boggling. We used to laugh at the sci-fi movies’ imagination of what the year 2000 would look like, but 2050? It will look like my favorite movies. It’s much closer than we think.
Two Truths and A Lie:
A fun little game is always best to help us think through and learn. I will have one of these in each newsletter, and the next one will have the answer.
Rules:
1. Guess which statement is true and which one is a lie that’s myths about NVIDIA.
2. The ratio of truth to lie is 2:1
Game NL#39:
The G in GPU is for graphics. The core AI chip we’re using today was designed for computer graphics.
NVIDIA chips powers all of AI applications today.
NVIDIA makes AI chips.
Which one is wrong?
Answers for this game will be in the next newsletter.
Answers for NL#38
Lie: #1 – You will not need to study computer science and engineering to upskill for AI. You will however need to understand certain basic concepts.
Remember: Don’t commit acts of AI Unwittingly, do it Wittingly…